For a service across multiple network domains, a protection mechanism may be deployed in each of the multiple network domains to implement, in a sectionalized-protection manner, end-to-end protection for the service. To prevent a service protection failure when a single node fails, a dual-node interconnection manner is generally used between network domains. In this case, different from a traditional linear protection mechanism with a single source and a single sink, multiple source nodes and/or multiple sink nodes exist in a single protection domain.
In the current multi-domain network protection mechanism, multiple protection paths need to be configured on a protection switching node to implement protection in a multi-source and/or multi-sink manner in a single network domain. Each protection path connects to a different source and a different sink, so that protection switching can be effectively implemented for a protected service when a multi-domain network fails. This mechanism requires variations in functions of the protection switching node. However, on a service access side, a service accesses a network through a single service access node, and the service access node generally supports only the traditional single-source single-sink linear protection mechanism. Because a large number of service access nodes exist on the access side, in the current protection mechanism, a large number of service access nodes need to be reconstructed and upgraded to support protection on multiple source nodes and/or sink nodes, which causes complex device implementation and high device costs.